REVIEW New Phytol.

(2000), 147, 3–12

Research review
Root dynamics and global change : seeking
an ecosystem perspective

R I C H A R D J. N O R B Y"*    R O B E R T B. J A C K S O N#
"Environmental Sciences Division, Oak Ridge National Laboratory, Oak Ridge,
TN 37831–6422, USA
#Department of Botany and Nicholas School of the Environment, Duke University,
Durham, NC 27708, USA
Received 24 March 2000 ; accepted 4 April 2000


Changes in the production and turnover of roots in forests and grasslands in response to rising atmospheric CO
#
concentrations, elevated temperatures, altered precipitation, or nitrogen deposition could be a key link between
plant responses and longer-term changes in soil organic matter and ecosystem carbon balance. Here we summarize
the experimental observations, ideas, and new hypotheses developed in this area in the rest of this volume. Three
central questions are posed. Do elevated atmospheric CO , nitrogen deposition, and climatic change alter the
#
dynamics of root production and mortality ? What are the consequences of root responses to plant physiological
processes ? What are the implications of root dynamics to soil microbial communities and the fate of carbon in soil ?
Ecosystem-level observations of root production and mortality in response to global change parameters are just
starting to emerge. The challenge to root biologists is to overcome the profound methodological and analytical
problems and assemble a more comprehensive data set with sufficient ancillary data that differences between
ecosystems can be explained. The assemblage of information reported herein on global patterns of root turnover,
basic root biology that controls responses to environmental variables, and new observations of root and associated
microbial responses to atmospheric and climatic change helps to sharpen our questions and stimulate new research
approaches. New hypotheses have been developed to explain why responses of root turnover might differ in
contrasting systems, how carbon allocation to roots is controlled, and how species differences in root chemistry
might explain the ultimate fate of carbon in soil. These hypotheses and the enthusiasm for pursuing them are based
on the firm belief that a deeper understanding of root dynamics is critical to describing the integrated response
of ecosystems to global change.

Key words : fine roots, root dynamics, turnover, climatic change, global change, elevated CO , nitrogen deposition,
#
mycorrhizas.

to changing water and temperature regimes, or the


fate of anthropogenic nitrogen compounds deposited
Tansley (1935) in describing the concept of the onto ecosystems, the key components of ecosystem
ecosystem commented on ‘ … the necessity for in- response may reside out of sight – the belowground
vestigation of all the components of the ecosystem system of roots, soil and associated microorganisms.
and of the ways they interact to bring about Changes in the production and turnover of roots
approximation to dynamic equilibrium. That is the in forests and grasslands in response to rising
prime task of the ecology of the future.’ With respect atmospheric CO concentrations, elevated tempera-
#
to the current mandate to scientists to unravel the tures, altered precipitation, or N deposition could be
complexity of terrestrial ecosystem responses to a key link between plant responses and longer-term
global change, the future is now. Whether the changes in soil organic matter and ecosystem C
current research (or policy) question concerns the balance. While Tansley’s call for integrated studies
capacity for forests and grasslands to sequester of all ecosystem components has long been em-
carbon as the atmosphere becomes progressively braced, ecosystem-level observations of root pro-
enriched with CO , or the resistance of communities duction and mortality in response to global change
#
*Author for correspondence (tel j1 865 576 5261 ; fax j1 865
variables are just starting to emerge. The papers in
576 9939 ; e-mail rjn!ornl.gov). this volume highlight the current state of knowledge
4 REVIEW R. J. Norby and R. B. Jackson

about root responses to atmospheric and climatic major biomes across the globe and analyzed the data
change and their interaction with ecosystem proper- set for broad-scale patterns along climatic gradients.
ties. Turnover rates of fine roots increased exponentially
This collection revolves around three central with mean annual temperature in forests and grass-
questions : lands, but surprisingly, there was no relationship
$ Do elevated atmospheric CO , N deposition, and with precipitation once temperature effects were
# accounted for. Does this result suggest that root
climatic change alter the dynamics of root pro-
duction and mortality ? turnover will increase in response to climatic
$ What are the consequences of root responses to warming ? Not necessarily. Gill & Jackson (2000)
plant physiological processes ? found that the global relationship did not predict the
$ What are the implications of root dynamics to soil relationship to interannual variability in climate at a
microbial communities and the fate of C in soil ? particular site.
Investigations of global change effects on root
These questions are considerably easier to ask than
dynamics should be based on a good understanding
they are to answer. Ecosystem-scale experiments on
of the internal and external controls of root pro-
global change phenomena are few and recent (or
duction and longevity. Without such a framework,
ongoing). Results are often ambiguous, contradic-
observations from a manipulative experiment are
tory and confounded by methodological problems.
strictly phenomenological and difficult to apply to
Although the body of knowledge may be immature
the broader and more complicated issue of global
for a definitive synthesis, we have an opportunity to
change. Eissenstat et al. (2000) reviewed compre-
highlight the critical uncertainties, generate new
hensive studies indicating fine-root longevity to be
hypotheses, and improve our techniques in current
related to a number of root characteristics such as
and future research programs.
diameter, tissue density, mycorrhizal infection, N
concentration, and basal respiration rates, analogous

to relationships that have been described for leaves
Before delving into the results of manipulative (Reich et al., 1997). These characteristics can be
experiments, it is useful to understand the scope of used to predict root longevity through an analysis of
the issues of root dynamics and the physiological and the costs of constructing and retaining roots, and this
ecological controls that may be sensitive to global can provide a framework for sorting out the net effect
change factors. The system of small, ephemeral roots on root systems of the myriad influences that
is an important part of terrestrial metabolism. atmospheric and climatic change could have on
Jackson et al. (1997) estimated that as much as 33% plants and the soil environment (Eissenstat et al.,
of global annual net primary productivity is used for 2000).
the production of fine roots, which have a relatively The internal controls on root deployment and
short life before they die and begin decomposing. A longevity require an understanding of C allocation
central concept in our discussion is that root turnover processes in plants. Farrar & Jones (2000) developed
is a key component of the C and nutrient cycling in a hypothesis of C allocation in plants that suggests
ecosystems and will probably be sensitive to many control is shared by roots and shoots, with two
global change factors (Gill & Jackson, 2000 ; Eissen- mechanisms underlying the distribution of control –
stat et al., 2000). As discussed further here, root regulation of phloem transport and control of gene
turnover is variously defined, but generally it is a expression by resource compounds (e.g. sugars and
measure of the production and mortality of roots N compounds). In discussing the implications of this
relative to the size of the standing crop of roots. A model to predictions of climatic change impacts,
plant that maintains roots longer (lower turnover) Farrar & Jones (2000) are careful to distinguish a
allocates less C to the production of new roots, but shorter-term response, which might simply be bigger
expends more energy (i.e. root respiration) in plants with larger roots systems and little change in
maintaining roots that may be less efficient at C partitioning, from longer-term responses that
nutrient uptake compared to a root newly deployed occur after the plants begin to deplete soil resources.
in a nutrient-rich microsite. When roots die and This scale-related consideration is especially im-
decompose, some of their C is released to the portant when attempting to extend the responses of
atmosphere and some may remain as soil organic individual plants to that of ecosystems (Norby et al.,
matter (SOM). Hence, root turnover is a major 1999), and must, therefore, be kept at the forefront
component of ecosystem C fluxes and the potential in all of our discussions.
of an ecosystem to sequester atmospheric C.
Clues to how global change factors might alter
  
root turnover, and therefore nutrient and C cycling,
come from an analysis of the variation in root Any discussion about root biology is inevitably
turnover across environmental gradients. Gill & intertwined with methodological issues. It is, of
Jackson (2000) assembled root turnover data for course, much more difficult to make in situ obser-
 REVIEW Root dynamics and global change 5

vations (qualitative or quantitative) of roots than of studies are somewhat ambiguous. Turnover was
aboveground portions of plants, and the relevance of defined by Gill & Jackson (2000) in their global
responses of roots extracted from their soil en- survey as annual belowground production divided
vironment is always suspect. As a result the data base by maximum belowground standing crop, following
on root responses to global change factors is much the definition employed by Dahlman & Kucera
smaller than that for aboveground processes, and (1965). Using this approach, an annual plant that
without standard methods for researchers to rely on, maintains its roots through the growing season
much of the data that does exist is confounded by would have a turnover equal to one, where maximum
methodological differences. Novel techniques are biomass is equal to belowground production. Other
being developed, but progress in understanding and researchers have based turnover on the mean or
quantifying root dynamics and function – especially minimum annual standing crop rather than the
with an ecosystem perspective – will lag until the maximum. Nadelhoffer (2000) defines turnover as
new techniques are widely used and accepted. annual root death divided by mean fine root biomass.
One particularly frustrating problem is the sub- Note that if annual net root productivity (annual
stantial confusion created by inconsistent use of increment) is zero, as might be expected in a fully
important terms, such as fine root and turnover. In developed perennial system or in an annual system in
considering trees and forests, it is important to which all roots die at the end of the season, then
differentiate the responses of relatively small, mortality is equal to production. Other researchers
ephemeral roots from those of larger, woody per- have used the term ‘ turnover ’ as annual root
ennial roots, and failure to do so creates critical production or mortality in grams per square meter
errors in scaling responses from seedlings to mature (e.g. DeLucia et al., 1999). Tingey et al. (2000) refer
trees (Norby, 1994). But there remains considerable to annual root mortality as absolute turnover in
variation within the category of ‘ small, ephemeral contrast to relative turnover, which they call turn-
roots ’, and the word ‘ fine ’ is variously used for all or over index. The distinction is especially important
parts of that category. The term fine root is when the standing crop of roots is changing from
meaningless with grasses – all of the roots are fine, year to year, as during stand development or under
although not all are ephemeral. the influence of a changing environment. If the
Perhaps more important than establishing a pre- primary research interest is in rates of C and N fluxes
cise definition of a fine root is the recognition that from roots to soil, then the absolute turnover rate
processes and responses vary not just with root (which could be called annual root mortality) may be
diameter but with other root characteristics as well. most useful, but for studies of root demography or to
Eissenstat et al. (2000) emphasize the importance of link with ecosystem models, then relative turnover
root order, that is, the position of a root within the (or simply ‘ turnover ’ as already defined) is the
branched hierarchy of the root system. Fine roots appropriate metric (Tingey et al., 2000).
with no daughter roots generally have higher N The quantification of root production and mor-
concentration, respiration rate, specific root length, tality is difficult and controversial. Researchers use
and a shorter life span than similar roots with several methods to determine root production and
branches. While the very small diameter roots are mortality, each with strengths and limitations.
generally thought to turn over most rapidly, and Historically, sequential harvesting of roots was the
many observations support this view, very fine roots most widely used method to determine root pro-
(especially of an earlier order in the branching duction. Although several sources of error hamper
sequence) may live a remarkably long time. An this method, numerous algorithms have been de-
innovative technique by Gaudinski et al. (2000 ; also veloped to increase the value of harvest data and new
pers. comm.) utilized "%C created during the early methods and technologies continue to be introduced
1960s from thermonuclear testing to date the C in that may alleviate some problems with harvest
fine roots collected from a forest. Though very fine methods (Singh et al., 1984 ; Burke & Raynal, 1994).
( 1 mm in diameter), some of these roots were By contrast, Nadelhoffer (2000) believes that
surprisingly long-lived, ranging from 2 to 16 yr. methods based on sequential root harvests are flawed
Clearly there is need for caution in associating fine and give the wrong conclusion about the effect of N
roots with fast turnover. Heterogeneity of roots availability on root turnover. He estimates root
increases the complexity of our investigations turnover by constraining it with an N budget, a
further. promising method but one that is sensitive to
The definition of root turnover has also created estimates of N mineralization rates (Lauenroth,
confusion in the literature and in discussion, which 2000). Minirhizotrons are in increasing use in global
can hinder effective communication. In a systems change studies, providing a method for frequent,
science approach, turnover is typically defined as nondestructive quantification of root length pro-
flux divided by pool size, or the proportion of the duction and mortality (Hendrick & Pregitzer, 1992).
root biomass that is produced or dies annually. They are not without problems, however. For an
Unfortunately, both the flux and the pool for root ecosystem C budget it is necessary to convert data on
6 REVIEW R. J. Norby and R. B. Jackson

root length production per minirhizotron to grams of roots and coarse roots is unnecessary. In an annual
C per square meter ; a method for doing this is crop system, the important research questions are
described by Tingey et al. (2000). Another problem more likely to revolve around root deployment and
can be the definition of root death, especially in resource capture rather than equilibrium responses
minirhizotron images where tactile cues and staining and C flux (Fitter et al., 1991). Root turnover is
techniques are not possible. Typically root dis- likely to be impacted by CO differently in annual
#
appearance is used instead, but this may include versus perennial plants (Pritchard & Rogers, 2000).
initial stages of decomposition (Tingey et al., 2000). In discussing the responses of crop roots to
The definition of root death is ambiguous under any elevated CO , Pritchard & Rogers (2000) emphasize
#
circumstances since death occurs gradually and part the importance of cell expansion and cell division
of a root can die and cease function while the and suggest that a key to understanding how root
remaining root still maintains important functions growth will change in a high-CO environment is to
#
such as transport (Comas et al., 2000). understand how carbohydrates, especially sucrose,
functions both as a substrate for growth and as a
     regulatory molecule. From this perspective, they
 review the literature and conclude that roots in a
high CO environment will be larger and more
Will root production, mortality, or turnover change #
highly branched, but less efficient in nutrient and
in response to increasing atmospheric CO con-
# water uptake (also see Berntson & Bazzaz, 1996 ;
centration, N deposition, or climatic warming and
Rogers et al., 1999). Root turnover is not an
changing patterns of precipitation ? We are aware
important issue with annual crop plants : Fitter et al.
that the number of field studies that can address this
(1996) found no changes in root turnover in CO -
question with the appropriate ecological focus is #
enriched wheat because there is little or no mortality
extremely limited. In asking the question, we are
of wheat roots until they all die synchronously at the
essentially highlighting the tremendous uncertainty
end of the life cycle.
in the interpretation of the meager data set, but in so
Arnone et al. (2000) surveyed the literature on root
doing, we may be outlining the highest priority
responses to CO in native grassland ecosystems.
research questions and key measurements for new #
Results have been inconsistent, with less than half of
experiments as they get started. The largest emphasis
the studies showing increased root growth in elevated
here is on the effects of elevated atmospheric CO ,
# CO . In their own study in a calcareous grassland in
because that is the global change factor on which #
Switzerland, Arnone et al. (2000) used minirhizo-
there have been the most manipulative experiments.
trons to test their hypothesis that increased root
Furthermore, we presume there is a direct feedback
production and mortality would explain the observed
between root turnover responses to increased CO
# increase in net ecosystem CO uptake in this system.
and the cycling of C through ecosystems and back to #
After 2 yr of CO enrichment, there were no
the atmosphere. Root dynamics in elevated CO has #
# differences in root production or mortality. There
therefore been identified by the Global Change and
was a shift, however, in the distribution of roots,
Terrestrial Ecosystems (GCTE) Project as an
with more being found in the upper soil layer in the
especially important issue for assessments of global
CO -enriched plots, which may have been a result of
change response (Canadell et al., 1999). #
increased soil moisture in elevated CO .
#
The scale of forests has made them more difficult
Responses to elevated CO
# to study in manipulative experiments, and in their
Root dynamics have been investigated in field CO synthesis of the responses of coniferous trees, Tingey
#
experiments in forests, grasslands, and agricultural et al. (2000) had to rely mainly on data from
systems. Fundamental differences in these eco- experiments with seedlings and saplings. Fine-root
systems and their root systems lead to distinct growth increased in elevated CO across a range of
#
research questions, as well as distinct methodological species and experimental conditions, but there was
issues. In forests (or in tree seedling studies meant to no clear indication that CO enrichment altered the
#
address forest issues), it is particularly important to proportion of C allocated to roots. For example, the
differentiate the responses of large woody roots, ratio of fine root to needle mass did not change in a
which may be important in C storage, from those of consistent pattern in these studies. If root mass or
fine roots, which are important in nutrient and water fine-root density increases in elevated CO simply
#
uptake and C flux (Norby, 1994). Forests are very because of a coordinated whole-plant growth re-
difficult to include in manipulative experiments, and sponse rather than a specific response of roots, we
the available data on root responses to elevated CO really have not learned anything new : bigger plants
#
in forests is very limited. Grassland systems, on the have bigger root systems. However, we cannot expect
other hand, can be studied as intact ecosystems in these observations of young, isolated plants to persist
manipulative experiments (e.g. Fitter et al., 1997), over time as trees develop into a forest. Eventually,
and the arbitrary separation of responses by fine we might expect fine-root density in a forest to reach
 REVIEW Root dynamics and global change 7

a maximum value, analogous to the aboveground turnover response, consistent with the tentative
canopy reaching a maximum leaf area index (LAI) conclusion of Pregitzer et al. (2000b), is that lower N
(although can a concept similar to ‘ canopy closure ’ availability in the pine stand (R. Oren, pers. comm.)
apply below ground ?) It then becomes especially limits the response of fine roots to elevated CO
#
important to focus on the dynamics of the root compared to that in the more fertile hardwood stand.
system – the longevity of roots and the turnover of A third possibility, and an unfortunate one, is that
root systems. The few observations of CO effects on the apparent differences in response results from the
#
root turnover in field experiments with conifers have different techniques for assessing root turnover
shown an increased rate of root loss, but the response (sequential harvest versus minirhizotron).
of relative turnover has been inconsistent (Tingey et Clearly, there have not been enough experiments
al., 2000). In the free-air CO enrichment (FACE) attempted of sufficient duration or spatial scale to
#
experiment in a 15-yr-old Pinus taeda stand in North permit conclusions about the effect of elevated
Carolina, USA, Allen et al. (2000) reported that fine- atmospheric CO on root dynamics. Root production
#
root production increased 37% in elevated CO , and and mortality have been increased by CO en-
# #
mortality plus decomposition (or absolute turnover) richment in some studies and not in others (but have
increased 26% (not statistically significant), but in not been seen to decrease). Differences in response
this system turnover rate relative to production did might be related to system-level properties or
not change. interactions with other environmental drivers, and
In the few CO -enrichment studies with deciduous new hypotheses are needed to guide new and ongoing
#
trees under field conditions, the roots have been experiments.
more responsive to CO enrichment than those of
#
conifers. In six studies, fine-root density (the mass of
Response to climatic change
roots per unit ground area) increased from 60 to
140% in elevated CO (Norby et al., 1999). Con- Atmospheric CO can influence root dynamics
# #
trasting with the conclusion of Tingey et al. (2000), through its direct interaction with C uptake and
fine-root density increased more than leaf area in allocation in plants, but the increasing concentration
every case, suggesting that the stimulation of fine- of CO in the atmosphere can also be expected to
#
root production was a specific response to elevated alter root dynamics indirectly through its role as a
CO . In addition to these static measures of root greenhouse gas affecting temperature and precipi-
#
response, fine-root production and mortality have tation patterns. As air temperature warms globally,
increased in CO -enriched deciduous trees in the few soil temperature can be expected to rise con-
#
studies that used minirhizotrons to measure root comitantly (Schlesinger & Andrews, 2000). Root
dynamics. Fine-root length production, mortality growth and mortality often increase with increasing
and net production all increased with CO en- temperature as long as soil moisture and nutrient
#
richment in an Acer saccharum–Acer rubrum as- availability are adequate (Pregitzer et al., 2000a),
semblage in open-top chambers in Oak Ridge, TN, conditions which must however be rare in natural
USA (E. G. O’Neill, pers. comm.). Elevated CO ecosystems. Generalizations can be difficult because
#
increased the diameter and length of individual roots different populations of plants have evolved to cope
of Populus tremuloides trees in open-top chambers with a wide range of air and soil temperature regimes.
(Pregitzer et al., 2000b). Fine-root length production Fitter et al. (1998), for example, concluded that root
and mortality and fine-root biomass all were in- growth in their grassland site was controlled by
creased by CO enrichment, but only in trees grown radiation flux, not by temperature. In addition, soil
#
in high-N soil ; there were no CO effects in low-N temperature has other effects, notably a change in N
#
soil. mineralization rate, and interacts with other en-
Fine-root turnover was studied in the deciduous vironmental factors to influence root functioning.
forest FACE experiment in a Liquidambar styraciflua Pregitzer et al. (2000a) also discuss the importance
stand in Tennessee using minirhizotrons (E. G. of seasonality in root dynamics, especially in per-
O’Neill, pers. comm.). In 2 yr of exposure to ennial plants, and speculate that global warming will
elevated CO , there has been no effect on annual net result in earlier root growth in the spring. Their
#
production (i.e. standing crop does not change), but analysis suggests that the flux of C from leaves to
both production and mortality are higher in CO - roots and into soil should increase with warming.
#
enriched stands, indicating increased turnover and Analyses of ecosystem response to warming often
increased input of C to the soil. These observations focus on the presumed increase in heterotrophic
contrast with the Pinus taeda FACE (Allen et al., respiration and the loss of C from the system
2000), perhaps because the deciduous forest has (Schlesinger & Andrews, 2000), and it is important
higher tree density and LAI and has roots that fully that the likelihood of a (partially) counterbalancing
occupy the soil, whereas the root system is still increase in C input by roots also be considered.
expanding in the pine stand. Another possible Interannual variation in precipitation clearly can
explanation for the apparent difference in root affect plant productivity, but the specific responses
8 REVIEW R. J. Norby and R. B. Jackson

of roots are not clear. At a global scale, precipitation initially taken up (directly by foliage or through a soil
did not explain patterns of fine-root turnover across pathway) is a critical one that should be further
different ecosystems once the effects of temperature investigated. Other global change factors, including
were accounted for, although turnover was related to CO , temperature and precipitation, which co-occur
#
precipitation at the local scale (Gill & Jackson, 2000). with increasing N deposition globally, are likely to
Joslin et al. (2000) conclude that responses of roots to modify root responses to increasing N availability.
irrigations treatments have been mixed and at any Various modeling approaches to exploring the
rate are difficult to extend to long-term change in interactive effects of CO and N deposition (dis-
#
precipitation at the ecosystem scale. They report on cussed in Norby, 1998) illustrate that the responses
the ‘ Throughfall Displacement Experiment ’ in to N deposition and CO enrichment are not likely to
#
which the long-term effects of both decreasing and be simple or additive. This is not surprising given
increasing water inputs to the forest floor of a mature the close linkages and feedbacks between the C and
deciduous forest in Tennessee are being investigated. N cycles in a plant, and roots are at a key intersection
After 5 yr of minirhizotron observations of fine roots between those cycles. Different assumptions about
in this experiment, clear evidence of significant allocation are particularly important to the net effect
changes at the stand level to drought has yet to of N deposition and CO enrichment (Norby, 1998),
#
emerge. Differences across treatments in net fine- which emphasizes the importance of developing a
root production were small and non-significant, better understanding of the control of C allocation to
although a trend of increasing net production (due to roots (Farrar & Jones, 2000).
less mortality) in the dry plot was noted. Perhaps the
most important finding of this study has been the
    
resilience of the forest in maintaining an active root
system with relatively constant fine-root mass de- While our focus here is on root dynamics at the
spite the interannual and imposed variation in soil ecosystem scale, there is an important interplay
moisture (Joslin et al., 2000). between root production, deployment, and turnover
and the physiological activity of roots, and this
interplay must be part of our analyses of the impacts
Nitrogen deposition
of global change. For example, increased nutrient
Terrestrial ecosystems around the world, but es- uptake might be observed in an isolated plant in
pecially in North America and Europe, are receiving elevated CO as a result of a larger plant having a
#
unprecedented amounts of N compounds, ammonia large root system. Scaling this observation to an
and nitrogen oxides. The primary route by which ecosystem, however, requires additional information
these compounds enter ecosystem cycles after being on whether there are ecosystem-level constraints to
deposited from the air is through the soil pools of the total size of the root system, coupled with
nitrate and ammonium. Therefore, understanding physiological information on the response of root-
how N deposition might influence root dynamics specific nutrient uptake efficiency. Some of the
comes largely from analyzing the response of roots to presumed controls (or correlates) of root turnover,
natural or applied gradients of N availability. including respiration rate or nutrient and carbo-
Different approaches to measuring fine roots have hydrate content, relate directly to the physiological
yielded different conclusions about the relationship activity of the root. The challenge of incorporating
between N availability and fine-root turnover in plant physiology into global change predictions is
forests, but Nadelhoffer (2000) argues that the best the trade-off between a more mechanistic basis to the
evidence indicates that with increasing N avail- predictions and picking which physiological attri-
ability, fine-root biomass typically decreases and butes scale to meaningful ecosystem-level changes.
turnover increases. He develops the hypothesis that A sustained growth response of plants to elevated
if N deposition increases N mineralization, nitrifi- CO in unmanaged ecosystems is expected to require
#
cation, or N availability to plants, then fine-root increased N availability because short-term increases
biomass will decrease but fine-root production and in nutrient-use efficiency cannot be sustained in-
turnover will increase. In forests at late stages of N definitely (Norby et al., 1986 ; Bazzaz, 1990 ; Lloyd &
saturation (Aber et al., 1985), the decreased root Farquhar, 1996). The capacity for a plant to take up
biomass will contribute to a loss of nitrate from the nutrients depends on both the size of the root system
system, and after prolonged and highly elevated N and its efficiency to deploy roots at the time and place
deposition, nutrient imbalances could disrupt fine- nutrients are present (Fitter et al., 1991), as well as
root functions. on the efficiency by which a particular root segment
A complete understanding of the interaction can take up a nutrient from the soil solution.
between N deposition and root dynamics depends on BassiriRad (2000) has analyzed the latter issue
many interacting factors, particularly the native through careful studies of the kinetics of NH + and
%
fertility of the soil. Gifford et al. (1996) concluded NO − uptake and the influence of CO enrichment,
$ #
that the assumption about how the exogenous N is soil temperature and N deposition. The response of
 REVIEW Root dynamics and global change 9

uptake kinetics as reported in the literature is highly whole-plant source–sink relationships. Clearly, re-
variable, but BassiriRad (2000) suggests that much sponses of roots to global change cannot be studied
of this variability may be attributable to experimental in isolation from the rest of the plant and the myriad
protocol (though there is insufficient evidence yet to influences of multiple, fluctuating, environmental
know), such as whether both high and low affinity factors.
transport systems were involved or only the more
ecologically relevant high affinity system. Never-
    
theless, there do appear to be species-specific
differences in response to elevated CO , and this In addition to their fundamental role in the growth
#
raises the possibility of shifts in community com- and physiology of plants, roots also are important
position resulting from CO inutrient interactions conduits by which C enters the soil. Root turnover is
#
(Berntson et al., 1998 ; BassiriRad, 2000). Increased a critical mediator of C cycling in ecosystems, and it
soil temperature generally increases N uptake ca- follows that the response of roots to global change
pacity, but the response seems to be greater in factors will be an important determinant of net
species from warm climates with fluctuating soil ecosystem production and the potential for eco-
temperatures than from those in cold climates systems to sequester fossil-derived C in long-term
(BassiriRad, 2000). pools.
Nutrient uptake is an energy-demanding process In nature, the roots of almost all plants are joined
that accounts for a significant fraction of root symbiotically with fungi to form the mycorrhizal
respiration. Atkin et al. (2000) considered how root partnership. The fungal partner resides at the
respiration rates might respond to climatic warming. interface between plant and soil and is a link in the
Root respiration is an important part of the C budget chain of transfers by which C moves from plant to
of ecosystems, although the C flux from root soil (Staddon et al., 1999). Hence, the effect of
respiration is often lumped with respiration by soil elevated CO or other global change factors (which
#
heterotrophs. Root respiration is very sensitive to are less studied) on the establishment, growth and
changes in temperature, yet it is very difficult to turnover of mycorrhizal fungi is an important factor
predict its response to climatic warming because it in assessment of C cycling in ecosystems. The
can acclimate rapidly to changes in growth tem- literature describing effects of CO enrichment on
#
perature, and the degree of acclimation varies among mycorrhizal colonization is inconclusive, largely
species (Atkin et al., 2000). Pregitzer et al. (2000a), because of the failure to separate the overall effect of
however, cite several examples where there was no CO on plant growth from any specific effect on
#
evidence of acclimation of root respiration to chang- mycorrhizas (a problem in many other aspects of
ing temperature. They suggest that the natural CO research). A more mechanistic understanding of
#
seasonal and diurnal fluctuations in soil temperatures how elevated CO could affect mycorrhizas must
#
that occur in temperate and boreal ecosystems might come from an improved understanding of C fluxes in
prevent significant acclimation in the field, in mycorrhizal plants under varying CO concentration
#
contrast to laboratory conditions with controlled, (Fitter et al., 2000). They argue that research on
constant temperature. Further complicating the mycorrhizas has been too ‘ plant centered ’ and to
extension to an ecosystem scale, respiration re- better understand mycorrhizal responses to global
sponses are likely to be confounded by any factor change there should be a new emphasis on the
(e.g. elevated CO or air temperature) that affects biology of the fungal partner – a ‘ mycocentric ’
#
shoot activity and C supply to roots (Atkin et al., approach. In particular, the diversity among fungal
2000). species in their responses to CO and the services
#
As discussed by Farrar & Jones (2000), carbo- they provide to the plant (e.g. P uptake) warrant
hydrate supply to roots is a key process linking more study.
environmental effects on shoot physiology to the Mycorrhizal tissue comprises a substantial fraction
changes in the growth and function of roots. Total of soil organic matter in many systems (Treseder &
nonstructural carbohydrate (TNC) can comprise Allen, 2000), and aspects of the biology and
from 4 to 23% of fine-root dry mass (Pregitzer et al., chemistry of mycorrhizal hyphae can influence the
2000a). In a simulation of root processes, Pregitzer cycling of C to soil. Although microarthropods graze
et al. (2000a) show that regardless of the TNC on live hyphae, they do not have a big impact on C
concentration Acer saccharum fine roots can live only cycling, which is determined by the turnover rate of
a short time without a continual supply of TNC. hyphae. Hyphal turnover is almost certainly greater
Respiratory acclimation to temperature is a critical than that of roots (Fitter et al., 2000), but an
factor in determining the lifespan of a root in relation important fraction of the hyphal mass is recalcitrant
to leaf activity and the supply of TNC. They compounds like chitin and glomalin that can account
conclude that during the growing season rates of for a significant pool of soil organic matter (Treseder
fine-root respiration and root longevity are likely to & Allen, 2000). Field studies (e.g. Rillig et al., 1999)
be closely linked to the rate of photosynthesis and have suggested effects of elevated CO on mycorr-
#
10 REVIEW R. J. Norby and R. B. Jackson

hizal community composition and the productivity $ Complications of uncontrolled, interacting en-
and decomposition of hyphae could lead to an vironmental variables.
increase in the amount of C sequestered in intact $ Difficulties in generalizing in the face of broad
hyphae and their residual components. Hyphal variability in responses among plants, fungal
responses to increases in N availability (as might partners and soil microbes, as well as the varia-
result from N deposition) have been inconsistent, bility introduced by methodology.
and the most general response may be shifts in
community composition based on differential sen- All of these issues, and many others, have been
sitivity to N. Because mycorrhizal groups differ in invoked in this set of papers. But at the same time
their growth rate and recalcitrant chitin content, new hypotheses have been introduced to suggest
these shifts could have implications for C immobili- why responses of root turnover might differ in
zation in SOM, but a great deal more work on contrasting systems, how C allocation to roots is
mycorrhizal physiology is needed before any pre- controlled, and how species differences in root
dictions can be made (Treseder & Allen, 2000). chemistry might explain the ultimate fate of C in
What then is the net result of the CO effects on soil.
#
fine-root productivity, turnover, and the flux of C Even if all the problems were resolved, our
through mycorrhizas on the fate of C in soil and assembled data would still be insufficient for ad-
sequestration in long-lived SOM pools ? Zak et al. dressing the larger issues in which we are all
(2000) attempted to address this question by sum- ultimately interested. That is, how will atmospheric
marizing data from 47 reports on soil C and N and climatic change alter the net productivity of
cycling under elevated CO , focusing on those pools ecosystems over decadal time scales ? Such far
#
and processes that are important control points for reaching questions can be addressed only with the
the belowground flow of C and the closely linked N use of simulation models. Ideally, those models are
fluxes. Soil and microbial respiration were almost informed by the results of experiments and the best
always more rapid under elevated CO because more compilation of our understanding of processes and
#
C entered the soil and the additional substrate was responses.
metabolized by soil microorganisms. This leaves as Woodward & Osborne (2000) summarize how
an open question the net effect of increased influx roots are currently considered in models used to
and increased efflux, and Zak et al. (2000) found that address global change issues (see also Jackson et al.,
microbial biomass, gross N mineralization, microbial 2000). The range of approaches used for simulating
immobilization and net N mineralization had such a root behavior in models has been fairly narrow, and
high degree of variability that there was no basis for modeling has not been as fruitful at providing new
any prediction about how microbial activity and ideas to explore experimentally for belowground
rates of soil C and N cycling will change with CO processes as for aboveground processes. The level of
#
enrichment. To remedy this fundamental gap in our detail currently used in models varies dramatically
understanding of ecosystem response to global depending on the model and whether the goal might
change, Zak et al. (2000) argue that we need to be to simulate water uptake for transpiration or
understand better how differences between plant the role of roots in soil C storage. Woodward &
taxa in fine-root production, mortality and bio- Osborne (2000) rightly point out that many global
chemistry may influence microbial metabolism and models perform well for today’s conditions with
lead to different patterns of soil C and N cycling. little or no representation of roots. One reason for
this agreement is that the major root activities of
:    resource capture are often correlated (e.g. uptake of
    N and transpired water). However, we do not know
if these correlations will stay the same in the future.
Most of the papers in this volume conclude that not Furthermore, extensive land use and land cover
enough is known to draw any firm conclusions about change are creating novel combinations of climate
how global change factors will impact root dynamics and plant life forms for which such relationships may
or how changes in root dynamics might affect plant not hold. The moderate success of all approaches
physiology or C cycling in soil. The problems are (including those that ignore roots completely), the
manifold : lack of appropriate global root data sets, and trade-
$ Daunting methodological problems in measuring offs between mechanistic detail and over-parameteri-
unseen roots without disturbing the system of zation all call into question how extensively root data
study. can and will be incorporated into global models in
$ An inability to sense belowground variables the future (Woodward & Osborne, 2000).
remotely at regional and global scales. Here we have a major challenge. Root biologists
$ Experimental challenges in simulating future accept on faith that a deeper understanding of roots
atmospheric and climatic conditions at a useful is critical to describing the integrated response of
spatial scale. ecosystems to global change, yet many global models
 REVIEW Root dynamics and global change 11

appear to work fairly well for today’s conditions with rising global CO levels. Annual Review of Ecology and
#
Systematics 21 : 167–196.
only rudimentary information on roots. Further- Berntson GM, Bazazz FA. 1996. The allometry of root
more, the broad patterns in root turnover observed production and loss in seedlings of Acer rubrum (Aceraceae) and
globally cannot necessarily be explained by detailed Betula papyrifera (Betulaceae) : implications for root dynamics
in elevated CO . American Journal of Botany 83 : 608–616.
local observations (Gill & Jackson, 2000). We cannot #
Berntson GM, Rajakaruna N, Bazzaz FA. 1998. Growth and
yet answer the questions put forth in the beginning nitrogen uptake in an experimental community of annuals
of this paper. There are insufficient data to support exposed to elevated atmospheric CO . Global Change Biology 4 :
#
any generalizations about effects of elevated CO on 607–626.
# Burke MK, Raynal DJ. 1994. Fine root growth phenology,
root turnover or the consequence of root turnover on production, and turnover in a northern hardwood forest
the C flux to long-lived soil C pools. The responses ecosystem. Plant and Soil 162 : 135–146.
to increased temperature are obscured by the Canadell J, Ingram J, Noble I, eds. 1999. Global Change
and Terrestrial Ecosystems Implementation Plan. IGBP Report
occurrence of acclimation, but this issue at least 47. Focus 1 : Ecosystem Physiology and Global Change. http:\\
appears to be tractable. The challenge, then, is to www.gcte-focus1.org\publications\Focus1Implementation.pdf
overcome the serious methodological and analytical Comas LH, Eissenstat DM, Lakso AN. 2000. Assessing root
problems and assemble a more comprehensive data death and root system dynamics in a study of grape canopy
pruning. New Phytologist 147 : 171–178.
set with sufficient ancillary data that differences Dahlman RC, Kucera CL. 1965. Root productivity and turnover
between ecosystems can be explained. Efforts to link in native prairie. Ecology 46 : 84–89.
root attributes mechanistically or correlatively with DeLucia EH, Hamilton JG, Naidu SL, Thomas RB, Andrews
JA, Finzi A, Lavine M, Matamala R, Mohan JE, Hendrey
remotely sensed variables are needed. An additional GR, Schlesinger WH. 1999. Net primary production of a
challenge is further to develop conceptual and forest ecosystem with experimental CO enrichment. Science
#
predictive models of C allocation in plants so that 284 : 1177–1179.
Eissenstat DM, Wells CE, Yanai RD, Whitbeck JL. 2000.
environmental influences on leaves can be translated Building roots in a changing environment : implications for
into the responses that follow in roots. New ideas root longevity. New Phytologist 147 : 33–42.
that might account for the large uncertainty in the Farrar JF, Jones DL. 2000. The control of carbon acquisition by
roots. New Phytologist 147 : 43–53.
ultimate fate of the C in roots need to be tested. And Fitter AH, Graves JD, Self GK, Brown TK, Bogie DS, Taylor
clear challenges remain : to integrate above- and K. 1998. Root production, turnover and respiration under two
belowground plant functioning and to demonstrate grassland types along an altitudinal gradient : influence of
that roots matter – that global predictions of temperature and solar radiation. Oecologia 114 : 20–30.
Fitter AH, Graves JD, Wolfenden J, Self GK, Brown TK,
terrestrial response to atmospheric and climatic Bogie D, Mansfield TA. 1997. Root production and turnover
change will be improved with a better charac- and carbon budgets of two contrasting grasslands under
terization of root dynamics. ambient and elevated atmospheric carbon dioxide concentra-
tions. New Phytologist 137 : 247–255.
Fitter AH, Heinemeyer A, Staddon PL. 2000. The impact
of elevated CO and global climate change on arbuscular
               #
mycorrhizas : a mycocentric approach. New Phytologist 147 :
179–187.
This paper was prepared with support from the Office of Fitter AH, Self GK, Wolfenden J, van Vuuren MMI, Brown
Biological and Environmental Research, US Department TK, Williamson L, Graves JD, Robinson D. 1996. Root
of Energy, at the Oak Ridge National Laboratory (ORNL). production and mortality under elevated atmospheric carbon
ORNL is managed by UT-Battelle, LLC, for the US dioxide. Plant and Soil 187 : 299–306.
Department of Energy under contract DE-AC05-00O- Fitter AH, Stickland TR, Harvey ML, Wilson GW. 1991.
Architectural analysis of plant–root systems. 1. Architectural
R22725. This work is part of the GCTE Core Project of
correlates of exploitation efficiency. New Phytologist 118 :
the International Geosphere–Biosphere Programme. Pub- 375–382.
lication No. 5000, Environmental Sciences Division, Oak Gaudinski JB, Trumbore SE, Davidson EA, Zheng S. 2000.
Ridge National Laboratory, Oak Ridge, TN. Soil carbon cycling in a temperate forest : radiocarbon-based
estimates of residence times, sequestration rates and par-
titioning of fluxes. Biogeochemistry. (In press.)
Gifford RM, Lutze JL, Barrett D. 1996. Global atmospheric
 change effects on terrestrial carbon sequestration : exploration
Aber JD, Melillo JM, Nadelhoffer KJ, McClaugherty C, with a global C- and N-cycle model (CQUESTN). Plant and
Pastor J. 1985. Fine root turnover in forest ecosystems in Soil 187 : 369–387.
relation to quantity and form of nitrogen availability. A Gill RA, Jackson RB. 2000. Global patterns of root turnover for
comparison of two methods. Oecologia 66 : 317–321. terrestrial ecosystems. New Phytologist 147 : 13–31.
Allen AS, Andrews JA, Finzi AC, Matamala R, Richter DD, Hendrick RL, Pregitzer KS. 1992. The demography of fine
Schelsinger WH. 2000. Effects of free-air CO enrichment roots in a northern hardwood forest. Ecology 73 : 1094–1104.
# Jackson RB, Mooney HA, Schulze ED. 1997. A global budget
(FACE) on belowground processes in a Pinus taeda forest.
Ecological Applications 10 : 437–448. for fine root biomass, surface area, and nutrient contents.
Arnone JA III, Zaller JG, Spehn E, Niklaus PA, Wells CE, Proceedings of the National Academy of Sciences, USA 94 :
Ko$ rner Ch. 2000. Dynamics of root systems in native 7362–7366.
grasslands : effects of elevated atmospheric CO . New Phy- Jackson RB, Schenk HJ, Jobba! gy EG, Canadell J, Colello GD,
# Dickinson RE, Field CB, Friedlingstein P, Heimann M,
tologist 147 : 73–85.
Atkin OK, Edwards EJ, Loveys BR. 2000. Response of root Hibbard K, Kicklighter DW, Kleidon A, Neilson RP,
respiration to changes in temperature and its relevance to global Parton WJ, Sala OE, Sykes MT. 2000. Belowground con-
warming. New Phytologist 147 : 141–154. sequences of vegetation change and their treatment in models.
BassiriRad H. 2000. Kinetics of nutrient uptake by roots : Ecological Applications 10 : 470–483.
responses to global change. New Phytologist 147 : 155–169. Joslin JD, Wolfe MH, Hanson PJ. 2000. Effects of altered water
Bazzaz FA. 1990. The response of natural ecosystems to the regimes on forest root systems. New Phytologist 147 : 117–129.
12 REVIEW R. J. Norby and R. B. Jackson

Lauenroth WK. 2000. Methods of estimating belowground net Reich PB, Walters MB, Ellsworth DS. 1997. From tropics to
primary production. In : Sala OE, Jackson RB, Mooney HA, tundra : global convergence in plant functioning. Proceedings of
Howarth R, eds. Methods in ecosystem science. New York, USA : the National Academy of Sciences, USA 94 : 13730–13734.
Springer, 58–71. Rillig MC, Wright SF, Allen MF, Field CB. 1999. Rise in
Lloyd J, Farquhar GD. 1996. The CO dependence of photo- carbon dioxide changes soil structure. Nature 400 : 628.
# Rogers HH, Runion GB, Prior SA, Torbert HA. 1999.
synthesis, plant growth responses to elevated atmospheric CO
# Responses of plants to elevated atmospheric CO : root growth,
concentrations and their interaction with soil nutrient status. I. #
General principles and forest ecosystems. Functional Ecology mineral nutrition, and soil carbon. In : Luo Y, Mooney HA,
10 : 4–32. eds. Carbon dioxide and environmental stress. San Diego, CA,
Nadelhoffer KJ. 2000. The potential effects of nitrogen de- USA : Academic Press, 215–244.
position on fine-root production in forest ecosystems. New Schlesinger WH, Andrews JA. 2000. Soil respiration and the
Phytologist 147 : 131–139. global carbon cycle. Biogeochemistry 48 : 7–20.
Norby RJ. 1994. Issues and perspectives for investigating root Singh JS, Lauenroth WK, Hunt HW, Swift DM. 1984. Bias and
responses to elevated atmospheric carbon dioxide. Plant and random errors in estimators of net root production : a simulation
Soil 165 : 9–20. approach. Ecology 65 : 1760–1764.
Staddon PL, Robinson D, Graves JD, Fitter AH. 1999. The
Norby RJ. 1998. Nitrogen deposition : a component of global
δ"$C signature of the external phase of a Glomus mycorrhizal
change analyses. New Phytologist 139 : 189–200.
fungus : determination and implications. Soil Biology and
Norby RJ, Pastor J, Melillo JM. 1986. Carbon–nitrogen
Biochemistry 31 : 1067–1070.
interactions in CO -enriched white oak : physiological and long-
# Tansley AG. 1935. The use and abuse of vegetational concepts
term perspectives. Tree Physiology 2 : 233–241. and terms. Ecology 16 : 284.
Norby RJ, Wullschleger SD, Gunderson CA, Johnson DW, Tingey DT, Phillips DL, Johnson MG. 2000. Elevated CO and
Ceulemans R. 1999. Tree responses to rising CO : impli- #
# conifer roots : effects on growth, life span and turnover. New
cations for the future forest. Plant, Cell & Environment 22 : Phytologist 147 : 87–103.
683–714. Treseder KK, Allen MF. 2000. Mycorrhizal fungi have a
Pregitzer KS, King JS, Burton AJ, Brown SS. 2000a. Responses potential role in soil carbon storage under elevated CO and
of tree fine roots to temperature. New Phytologist 147 : 105–115. nitrogen deposition. New Phytologist 147 : 189–200.
#
Pregitzer KS, Zak DR, Maziasz J, DeForest J, Curtis PS, Woodward FI, Osborne CP. 2000. The representation of root
Lussenhop J. 2000b. Fine root growth, mortality, and processes in models addressing the responses of vegetation to
morphology in a factorial elevated atmospheric CO isoil N global change. New Phytologist 147 : 223–232.
#
availability experiment. Ecological Applications 10 : 18–33. Zak DR, Pregitzer KS, King JS, Holmes WE. 2000. Elevated
Pritchard SG, Rogers HH. 2000. Spatial and temporal de- atmospheric CO , fine roots and the response of soil micro-
ployment of crop roots in CO -enriched environments. New #
organisms : a review and hypothesis. New Phytologist 147 :
#
Phytologist 147 : 55–71. 201–222.